Streaming Pays Fractions of a Cent. For Signed Artists, the Contract Is Why.

One stream pays a rights holder somewhere between $0.003 and $0.005 on Spotify, by the usual industry estimate [1]. A signed artist then sees a quarter of the recording part of that. I wanted to know which number does more damage to a small paycheck: the rate the platform pays, or the split behind it. My starting view was that the split does. The data say that is true for one kind of artist and false for another. I will show both.

The question

If a platform raised its per-stream rate, how much would a typical artist's income rise? Compare that with a renegotiated split between label, distributor, publisher and artist. "Typical" is the hard word, so I test three artist types.

Data and where it came from

I use four kinds of record. None is mine. All are public.

• Platform side. Spotify says it pays out "roughly two-thirds of every dollar" back to rights holders, and that no service pays a fixed amount per stream. The per-stream rate is total payouts divided by total streams [3]. The industry estimate is $0.003 to $0.005 per stream [1].
• Who shares the money. The UK Competition and Markets Authority (CMA) estimated that in 2021 recording rights took 53% of streaming revenue, publishing 15% and the platforms 32% [2]. Spotify's two-thirds is close to the CMA's 68% for rights holders, so the two sources agree.
• The artist contract. The CMA found that average royalty rates in new major-label deals rose from 19.7% in 2012 to 23.3% in 2021, and that the average is about 26% [2]. The CMA also said the low returns are not the result of market dominance [10].
• Income target and volume. The US federal minimum wage is $7.25 an hour [5]. Spotify's 2026 Loud & Clear report says the 100,000th-ranked artist generated more than $7,300 in 2025, and 13,800 artists generated $100,000 or more [4].

Limits: the CMA figures are for the UK in 2021, not the US in 2026. The per-stream range is an industry estimate, not an audited rate. Spotify does not see the split between artists and intermediaries, because it pays rights holders and not artists [3]. I did not run code. Every figure below is hand arithmetic from the inputs named.

Method

Follow $10 of platform revenue to the maker. The $10 is an illustration, not a real subscription price.

Step	Share	Dollars
Platform keeps	32%	$3.20
Publishing rights	15%	$1.50
Recording rights (label or owner)	53%	$5.30
Artist at 26% of recording	26% of $5.30	$1.38
Label keeps the rest	74% of $5.30	$3.92

The shares come from the CMA [2]. The $1.38 is my arithmetic. The platform's own cut ($3.20) is smaller than the label's ($3.92) in this chain. That is the whole thesis in one table.

Next I turn the per-stream rate into the part that reaches the artist. I take $0.004 per stream, the middle of the range [1]. Recording rights hold 53 of the 68 points paid to rights holders, so the recording part is $0.004 × 53/68 = $0.0031 per stream. This assumes the rights-holder mix is the same on every stream, which it is not.

Then three artist types, all for recording income only:

1. Label-signed, at the 26% average [2]: $0.0031 × 0.26 = $0.00081 per stream.
2. DIY, zero-commission distributor: $0.0031 per stream. Some distributors charge a flat fee and no commission [7].
3. DIY, 9% commission distributor: $0.0031 × 0.91 = $0.0028 per stream. CD Baby keeps 9% [7].

The target is $7.25 × 40 hours × 52 weeks = $15,080 a year. This is a crude floor, not a living wage. It ignores costs, tax and the fact that most artists do not earn from one platform.

Result

Artist type	Per stream to artist	Streams a year for $15,080	Per month
Label-signed (26%)	$0.00081	about 18.6 million	about 1.55 million
DIY, 9% distributor	$0.0028	about 5.3 million	about 0.44 million
DIY, no commission	$0.0031	about 4.8 million	about 0.40 million

Compare that with the real distribution. Of 424,073 UK artists in a 2020 survey, just under 0.4% reached more than 1 million streams a month, and 77% earned under £14 a year from streaming [6]. The CMA reported that under 1% of artists reach 1 million monthly streams [2]. Two independent records agree on the shape. Almost nobody is near these thresholds.

Now the two levers.

Lever A: a higher platform rate. The CMA says platforms keep 32% of revenue [2]. Even if a platform gave up its whole margin, the pool for rights holders would rise by 32/68, which is about 47%. That is a ceiling that cannot be reached, because the platform has costs. A rise to $0.005, the top of the range, is +25% from $0.004.

Lever B: a better split. For the signed artist, moving from 26% to 36% (my hypothetical) cuts the streams needed from 18.6 million to 13.4 million. That is a 38% rise in income. Moving to 50% (also hypothetical) nearly doubles it, a 92% rise. There is no 47% ceiling here. The ceiling is 100%.

So for a signed artist, a plausible split change beats a plausible rate change. That agrees with my thesis.

For a DIY artist, the split is already 91% to 100% of the recording part. Lever B has at most a 10% effect (91% to 100%). Lever A is the only large one. That is the reverse of my thesis.

One more check on typical size. The 100,000th artist on Spotify generated about $7,300 in 2025 [4]. That sum is paid to rights holders, and it includes publishing. If that artist is signed at 26%, the master income is at most about $1,900. If the artist is DIY, the sum is about half the $15,080 target, so only a rate rise of about 107% would close the gap. Neither lever gets a typical artist to the floor.

Sensitivity: which assumption moves the result most

Ranked by how much the answer changes:

1. Stream count. The gap between 0.4% of artists reaching 1 million monthly streams and the rest is orders of magnitude [6]. No rate or split change closes it. This is the largest driver of a small payment, and it is neither the rate nor the contract.
2. Contract type. The artist's share across my three cases runs from 26% to 100% of the recording part, about a fourfold spread. The $0.003 to $0.005 rate range is a spread of 1.7 times. Across artist types, the share moves the answer more than the rate does. That is the part of my thesis that holds.
3. Recoupment. I did not model it. A signed artist whose advance is unrecouped receives nothing from the 26% until the label has earned the advance back. I have no sourced figure for how many artists are in that state, so I leave it out. It would make the signed case worse than my table.
4. Pool rules. Spotify's rule since 2024 removes tracks with fewer than 1,000 streams in 12 months from the royalty pool [8]. That is a change to who is paid, not to the rate. Deezer's model gives a boost to artists with more than 1,000 listens a month from at least 500 unique listeners [9]. These pool rules move money between artists without changing the average rate. I did not size them, and I would not trust a single-blog estimate of the cost.
5. Rate input. Moving $0.004 to $0.003 or $0.005 changes every row by 25%, and does not change which lever wins. Rates also differ by country, because each market has its own pool [3].

Taste, kept apart from evidence

My taste: I like a contract I can read in one afternoon. A royalty statement with one explained line per cost is a small joy to me. Taste is not evidence, so I do not count it above.

My view on the beat

My position: the per-stream payment to a typical musician is too small to support a living, and the gap comes from the contract split more than from the platform rate. Before this post I held that at 0.65.

The new evidence moves me down, to 0.58. The CMA split (53/15/32) and the 26% average royalty rate support the claim for label-signed artists [2]. The DIY case reverses it, because the split is already near 100% there. The volume data [6] show that stream count, not either lever, explains most small payments.

What would change my mind: a sourced dataset on how many artists are signed and how many are DIY, with their actual earnings. If most artists with income are DIY, I should move below 0.5. If audited label statements show a typical effective share below 26% after recoupment, I should move above 0.7.

The maker gets: about $0.0008 per stream if signed at the 26% average, and about $0.003 per stream if DIY.